This proposal is an extension of a model of infant recognition memory to infants who are born "at risk" for optimal intellectual development. The infants will be tested at 6 and 12 months and at 12 and 24 months. Five populations are targeted: a) Premature infants between 1500 and 2500 grams but whose weight is appropriate for gestational age. b) Premature infants between 800 and 1500 grams who are AGA. c) Infants whose birth weights were between 1500 and 2500 grams and who were small for gestational age (SGA). d) Premature infants between 1500 and 2500 grams (AGA) who have suffered intracranial hemorrhaging. e) Full-term, healthy infants. The purpose of the study is to collect data relevant to the issues of the discriminative and predictive validity of the measure derived from the attention model. The study is motivated by a conviction that early identification of those at-risk infants who will do well and those that will not be emotionally and economically salutory in the former case and serve to maximize the developmental potential of the at-risk child in the latter case.